Magnetic Resonance Imaging Grading Systems for Central Canal and Neural Foraminal Stenoses of the Lumbar and Cervical Spines With a Focus on the Lee Grading System

Overview of Cervical Spine Injuries Caused by Diving Into Shallow Water on Jeju Island: A 9-Year Retrospective Study in a Regional Trauma Center

Objective

Shallow water diving-related spinal cord injuries (SCIs) are a significant cause of cervical spine trauma, particularly in younger individuals. This study retrospectively evaluated the outcomes of patients with SCI caused by shallow-water diving accidents at a regional trauma center on Jeju Island, South Korea. The primary aim of this study was to investigate the relationships between the timing of treatment, injury characteristics, and prognosis.

Methods

A retrospective analysis was conducted of patients with cervical SCI resulting from shallow-water diving injuries admitted to the trauma center over a 9-year period. The data were obtained from medical records and neurological outcomes were measured using the American Spinal Injury Association scale. Statistical analyses, including correlation and multiple regression analyses, were performed to identify factors influencing prognosis.

Results

Thirty-four patients with cervical SCI resulting from shallow-water diving were included in this study. No statistically significant correlation was found between surgical timing and prognosis; however, significant correlations with prognosis were identified for mean canal compromise (MCC), mean spinal cord compression, and lesion length. In the multiple regression analysis, higher MCC and severe SCI, particularly hemorrhagic injury, were associated with prognosis. The mean time from injury to surgery was 1.25 days.

Conclusion

This study indicates that, specifically for patients with a higher MCC but less severe SCI, appropriate and more rapid intervention may improve prognosis. However, further large-scale studies are required to clarify the favorable factors and their role in achieving a good prognosis.

MRI evaluation of lumbar foraminal stenosis: correlation between a new quantitative evaluation and the qualitative Lee's classification

PURPOSE

Lumbar foraminal stenosis (LFS) accounts 8-11% of leg radiculopathy cases, typically assessed by MRI with qualitative classifications and more recently, quantitative ones. However, a combined surface area assessment of the foramen and exiting nerve root is lacking. We aimed to correlate the Nerve root/Foramen ratio (N/F ratio), a novel quantitative assessment for LFS, with Lee's classification.

METHODS

We studied lumbar spine MRI images of 36 patients eligible for degenerative lumbar spine surgery. Inclusion criteria comprised T2-weighted 3D images (120-210 slices) from L1 to S1 without artifacts. LFS severity was assessed using the N/F ratio and Lee's classification. Image analysis utilized sagittal reconstructions of T2-weighted 3D axial sequences using 3D slicer software. Quantitative analysis of 360 foramens was conducted using manual segmentation to calculate the N/F ratio from the cross-sectional areas of the foramen and exiting nerve root (higher score indicates more severe LFS). Qualitative LFS analysis was based on Lee's classification (grade 0 to 3).

RESULTS

LFS severity (N/F ratio and Lee's grade) increased in the craniocaudal direction. Stratified by level and side, the N/F ratio was significantly correlated with Lee's grade, except for the right L2L3 foramen. Applying a linear mixed model, a positive and significant association was found between Lee's classification and the N/F ratio (β = 0.10 [95%CI: 0.09 ; 0.10]; p < 0.001).

CONCLUSION

The N/F ratio emerges as an objective quantitative measure for LFS correlated to Lee's classification, integrating evaluation of perineural intraforaminal elements and exiting nerve root. Future automating manual segmentation could facilitate rapid LFS evaluation in daily clinic.

Quantification and classification of lumbar disc herniation on axial magnetic resonance images using deep learning models

PURPOSE

Application of a deep learning model visualization plugin for rapid and accurate automatic quantification and classification of lumbar disc herniation (LDH) types on axial T2-weighted MRIs.

METHODS

Retrospective analysis of 2500 patients, with the training set comprising data from 2120 patients (25,554 images), an internal test set covering data from 80 patients (784 images), and an external test set including data from 300 patients (3285 images). To enhance implementation, this study categorized normal and bulging discs as a grade without significant abnormalities, defining the region and severity grades of LDH based on the relationship between the disc and the spinal canal. The automated detection training and validation process employed the YOLOv8 object detection model for target area localization, the YOLOv8-seg segmentation model for disc recognition, and the YOLOv8-pose keypoint detection model for positioning. Finally, the stability of the detection results was verified using metrics such as Intersection over Union (IoU), mean error (ME), precision (P), F1 score (F1), Kappa coefficient (kappa), and 95% confidence interval (95%CI).

RESULTS

The segmentation model achieved an mAP50:95 of 98.12% and an IoU of 98.36% in the training set, while the keypoint detection model achieved an mAP50:95 of 93.58% with a mean error (ME) of 0.208 mm. For the internal and external test sets, the segmentation model's IoU was 97.58 and 97.49%, respectively, while the keypoint model's ME was 0.219 mm and 0.221 mm, respectively. In the quantification validation of the extent of LDH, P, F1, and kappa were measured. For LDH classification (18 categories), the internal and external test sets showed P = 81.21% and 74.50%, F1 = 81.26% and 74.42%, and kappa = 0.75 (95%CI 0.68, 0.82, p = 0.00) and 0.69 (95%CI 0.65, 0.73, p = 0.00), respectively. For the severity grades of LDH (four categories), the internal and external test sets showed P = 92.51% and 90.07%, F1 = 92.36% and 89.66%, and kappa = 0.88 (95%CI 0.80, 0.96, p = 0.00) and 0.85 (95%CI 0.81, 0.89, p = 0.00), respectively. For the regions of LDH (eight categories), the internal and external test sets showed P = 83.34% and 77.87%, F1 = 83.85% and 78.21%, and kappa = 0.77 (95%CI 0.70, 0.85, p = 0.00) and 0.71 (95%CI 0.67, 0.75, p = 0.00), respectively.

CONCLUSION

The automated aided diagnostic model achieved high performance in detecting and classifying LDH and demonstrated substantial consistency with expert classification.

Preprocedural Imaging Review Before Performing Epidural Steroid Injections: Analysis of Physician Practice Parameters

Introduction: Epidural steroid injections (ESIs) are a common interventional treatment for managing spinal pain complaints. Despite their widespread use, practice patterns among physicians performing ESIs vary significantly. This study aimed to evaluate preprocedural imaging review by pain physicians who perform ESIs in the cervical, thoracic, and lumbar spine. Methods: A survey was distributed to a cohort of physicians who regularly perform ESIs. The survey comprised questions regarding preprocedural imaging review before performing ESIs in the cervical, thoracic, and lumbar spine. The respondents included a diverse group of pain management physicians from various specialties and practice settings. Results: The results revealed that the majority of interventional pain management physicians personally interpret their own imaging, followed by a significant percentage of physicians who rely on the radiology reports. There were no physicians who did not perform any imaging review prior to ESIs. Whereas all respondents reported some form of imaging review, only 63.86%, 53.75%, and 64.44% reviewed the actual images prior to cervical, thoracic, and lumbar access, respectively. Conclusions: This survey provides initial data regarding imaging reviews among physicians who perform ESIs. Our results demonstrate that physicians treat imaging review as an essential component of the preprocedural process for performing ESIs, as all physicians reported that they perform some form of imaging review before performing ESIs. However, there is only partial adherence to the multidisciplinary working group opinion that segmental imaging should be reviewed for adequacy of space prior to cervical epidural access.

Artificial intelligence for segmentation and classification in lumbar spinal stenosis: an overview of current methods

PURPOSE

Lumbar spinal stenosis (LSS) is a frequently occurring condition defined by narrowing of the spinal or nerve root canal due to degenerative changes. Physicians use MRI scans to determine the severity of stenosis, occasionally complementing it with X-ray or CT scans during the diagnostic work-up. However, manual grading of stenosis is time-consuming and induces inter-reader variability as a standardized grading system is lacking. Machine Learning (ML) has the potential to aid physicians in this process by automating segmentation and classification of LSS. However, it is unclear what models currently exist to perform these tasks.

METHODS

A systematic review of literature was performed by searching the Cochrane Library, Embase, Emcare, PubMed, and Web of Science databases for studies describing an ML-based algorithm to perform segmentation or classification of the lumbar spine for LSS. Risk of bias was assessed through an adjusted version of the Newcastle-Ottawa Quality Assessment Scale that was more applicable to ML studies. Qualitative analyses were performed based on type of algorithm (conventional ML or Deep Learning (DL)) and task (segmentation or classification).

RESULTS

A total of 27 articles were included of which nine on segmentation, 16 on classification and 2 on both tasks. The majority of studies focused on algorithms for MRI analysis. There was wide variety among the outcome measures used to express model performance. Overall, ML algorithms are able to perform segmentation and classification tasks excellently. DL methods tend to demonstrate better performance than conventional ML models. For segmentation the best performing DL models were U-Net based. For classification U-Net and unspecified CNNs powered the models that performed the best for the majority of outcome metrics. The number of models with external validation was limited.

CONCLUSION

DL models achieve excellent performance for segmentation and classification tasks for LSS, outperforming conventional ML algorithms. However, comparisons between studies are challenging due to the variety in outcome measures and test datasets. Future studies should focus on the segmentation task using DL models and utilize a standardized set of outcome measures and publicly available test dataset to express model performance. In addition, these models need to be externally validated to assess generalizability.

Effect of intervertebral foramen area and width on postoperative pain relief in patients with cervical spondylotic radiculopathy

OBJECTIVE

This study aims to investigate the relationship between preoperative cervical intervertebral foramen width and area and the persistence of postoperative pain in patients diagnosed with cervical spondylotic radiculopathy (CSR).

METHODS

Patients were divided into two groups, based on their pain relief at the 6-month postoperative follow-up: the pain relief group and the persistent pain group. We compared various parameters, including age, sex, body mass index (BMI), duration of symptoms, preoperative Japanese Orthopedic Association (JOA) score, Neck Disability Index (NDI) score, postoperative ratio of disc space distraction, preoperative width of the intervertebral foramen (WIVF), and area of the intervertebral foramen (AIVF) between the two groups. Binomial logistic regression analysis was conducted to identify the factors affecting pain relief.

RESULTS

Significant differences were observed in preoperative WIVF, AIVF, duration of symptoms, preoperative NDI scores, and the ratio of disc space distraction between the two groups (all P < 0.05). Regression models indicated that symptom duration, preoperative NDI score and ratio of disc space distraction were negatively associated with pain relief, whereas preoperative WIVF and AIVF were positively associated with pain relief.

CONCLUSION

Preoperative WIVF and AIVF may be linked to persistent postoperative pain in patients with CSR.

Construct validity and reliability of the 2-minute step test in patients with lumbar spinal stenosis

AIM

This study aimed to investigate the validity and reliability of the 2-Minute Step Test (2MST) in patients with lumbar spinal stenosis (LSS).

METHOD

This cross-sectional study involved 55 volunteers, aged 18-80, diagnosed with lumbar spinal stenosis. The participants were assessed using the 2MST, the 2-Minute Walk Test (2MWT), the 6-Minute Walk Test (6MWT), the Visual Analogue Scale (VAS), and the Oswestry Disability Index (ODI). To evaluate the reliability of the 2MST, test-retest reliability was determined by administering the 2MST again after a 7-day interval. The reliability analysis included the intraclass correlation coefficient (ICC) with a 95% confidence interval (CI), standard error of measurement (SEM), and minimum detectable change at the 95% confidence level (MDC95). For concurrent validity, the correlations of the 2MST with the 2MWT, 6MWT, and ODI were assessed.

RESULTS

The 2MST exhibited excellent test-retest reliability (ICC = 0.94, SEM = 5.56, MDC95=15.41) in LSS patients. Additionally, the 2MST showed significant correlations with the 2MWT (r = 0.842, p < 0.001), 6MWT (r = 0.819, p < 0.001), and ODI score (r = -0.722, p < 0.001).

CONCLUSION

This study demonstrated that the 2MST is a valid and reliable tool for assessing exercise capacity in individuals with LSS.

CLINICALTRIALS.GOV ID

NCT06060821.

Conceptual review of outcome metrics and measures used in clinical evaluation of artificial intelligence in radiology

Artificial intelligence (AI) has numerous applications in radiology. Clinical research studies to evaluate the AI models are also diverse. Consequently, diverse outcome metrics and measures are employed in the clinical evaluation of AI, presenting a challenge for clinical radiologists. This review aims to provide conceptually intuitive explanations of the outcome metrics and measures that are most frequently used in clinical research, specifically tailored for clinicians. While we briefly discuss performance metrics for AI models in binary classification, detection, or segmentation tasks, our primary focus is on less frequently addressed topics in published literature. These include metrics and measures for evaluating multiclass classification; those for evaluating generative AI models, such as models used in image generation or modification and large language models; and outcome measures beyond performance metrics, including patient-centered outcome measures. Our explanations aim to guide clinicians in the appropriate use of these metrics and measures.

Relationships between stenosis severity, functional limitation, pain, and quality of life in patients with cervical spondylotic radiculopathy

Background/aim

This study aimed to examine the relationships between severity of stenosis, pain, functional limitation, disability, and quality of life in patients with cervical spondylotic radiculopathy.

Materials and methods

Patients (45 female, 19 male) with radiculopathy due to spondylotic changes in the cervical spine were included in this study. Stenosis severity (thecal sac cross-sectional area (CSA)), numbness, neck and arm pain severity, functional limitation (Cervical Radiculopathy Impact Scale), disability, and quality of life (EQ-5D-3L General Quality of Life Scale) were evaluated. The study was registered at ClinicalTrials.gov as NCT06001359.

Results

According to CSA values, 28 (43.75%) patients had severe stenosis and 36 (56.25%) had moderate stenosis, and the average CSA was 81.65 ± 10.08 mm2. Positive correlations were found between both neck and arm pain and neck disability (r = 0.597, r = 0.359), and negative correlations were found for the General Quality of Life Scale index score and EQ-5D-3L visual analog scale (r = -0.787, r = -0.518). There were significant positive correlations between Cervical Radiculopathy Impact Scale subscales and severity of stenosis, neck and arm pain, numbness, and disability (p < 0.05 for all). A significant negative correlation was observed between Cervical Radiculopathy Impact Scale subscales and quality of life (p < 0.01). Stenosis severity was correlated with pain, neck disability, and quality of life (p < 0.01 for all).

Conclusion

There are direct relationships between cervical spondylotic radiculopathy and neck and arm pain, numbness, disability, and quality of life. Additionally, an increase in the severity of cervical stenosis is associated with an increase in pain and disability.

Morphometric analysis of cervical neuroforaminal dimensions from C2-T1 using computed tomography of 1,000 patients

BACKGROUND

Race and sex differences are not consistently reported in the literature. Fundamentally, anatomical differences of cervical neuroforaminal dimensions (CNFD) amongst these groups would be important to know.

PURPOSE

To establish normative radiographic morphometric measurements of CNFD and uncover the influence of patient sex, race, and ethnicity while also considering anthropometric characteristics.

STUDY DESIGN

Retrospective radiographic morphometric study.

PATIENT SAMPLE

A total of 1,000 patients between 18 and 35 years of age who were free of spinal pathology.

OUTCOME MEASURES

Foraminal height, axial width, and area of cervical neural foramen.

METHODS

Cervical CTs were reviewed to measure CNFD, defined as follows: foraminal height, axial width, and area. Statistical analyses were performed to assess associations between CNFD, and patient height, weight, sex, race, and ethnicity.

RESULTS

CNFD measurements followed a bimodal distribution pattern moving caudally from C2-T1. Irrespective of disc level, cervical CNFD were as follows: left and right widths of 6.6±1.5 and 6.6±1.5 mm, heights of 9.4±2.4 and 9.4±3.2 mm, and areas of 60.0±19.5 and 60.6±20.7 mm2. Left and right foraminal width were highest at C2-C3 and lowest at C3-C4. Left and right foraminal height were highest at C7-T1 and C6-C7, respectively and lowest at C3-C4. Left and right foraminal areas were highest at C2-C3 and lowest at C3-C4. Significant differences were observed for all CNFD measurements across disc levels. CNFD did not vary based on laterality. Significant CNFD differences were observed with respect to patient sex, race, and ethnicity. Male height and area were larger compared to females. In contrast, female foraminal width was larger compared to males. The Asian cohort demonstrated the largest foraminal widths. White and Hispanic patients demonstrated the largest foraminal heights and areas. Black patients demonstrated the smallest foraminal widths, heights, and areas. Patient height and weight were only weakly correlated with CNFD measurements across all levels from C2-T1.

CONCLUSIONS

This study describes 36,000 normative measurements of 12,000 foramina from C2-T1. CNFD measurements vary based on disc level, but not laterality. Contrasting left- versus right-sided neuroforamina of the same level may aid in determining the presence of unilateral stenosis. Patient sex, race, and ethnicity are associated with CNFD, while patient anthropometric factors are weakly correlated with CNFD.

Cervical Column and Cord and Column Responses in Whiplash With Stenosis: A Finite Element Modeling Study

Spine degeneration is a normal aging process. It may lead to stenotic spines that may have implications for pain and quality of life. The diagnosis is based on clinical symptomatology and imaging. Magnetic resonance images often reveal the nature and degree of stenosis of the spine. Stenosis is concerning to clinicians and patients because of the decreased space in the spinal canal and potential for elevated risk of cord and/or osteoligamentous spinal column injuries. Numerous finite element models of the cervical spine have been developed to study the biomechanics of the osteoligamentous column such as range of motion and vertebral stress; however, spinal cord modeling is often ignored. The objective of this study was to determine the external column and internal cord and disc responses of stenotic spines using finite element modeling. A validated model of the subaxial spinal column was used. The osteoligamentous column was modified to include the spinal cord. Mild, moderate, and severe degrees of stenosis commonly identified in civilian populations were simulated at C5-C6. The column-cord model was subjected to postero-anterior acceleration at T1. The range of motion, disc pressure, and cord stress-strain were obtained at the index and superior and inferior adjacent levels of the stenosis. The external metric representing the segmental motion was insensitive while the intrinsic disc and cord variables were more sensitive, and the index level was more affected by stenosis. These findings may influence surgical planning and patient education in personalized medicine.

Value of chemical shift imaging in the evaluation of neural foramen stenosis

Background

Chemical shift Magnetic Resonance Imaging (MRI) is often routinely acquired to assess a spectrum of spinal lesions due to its ability versatility to obtain rapid sequences at the expense of spatial resolution images. It is one of the quickest sequences to acquire at the expense of spatial resolution.

Objective

In this study we assess the diagnostic efficacy of Chemical shift Magnetic Resonance Imaging (MRI) in the evaluation of Neural Foraminal stenosis.

Materials and methods

Conventional T2, T1 and STIR sagittal and axial images as well as 'in' and 'out' phase chemical shift sagittal MRI sequences of 25 consecutive patients presenting with back pain and radiculopathy were reviewed. The degree of neural stenosis in the lumbar spine foramina on both sides was graded using the Lee classification on T2 and 'in' and 'out' phase MRI sequences by two independent MSK radiologists. Statistical analysis was performed using paired T-Test and Cohen's weighted kappa test was applied as a measure of agreement between the two observed raters.

Results

A total of 250 lumbar neural foramina were assessed. There was substantial agreement (Cohen's weighted kappa) for both raters between 'in' and 'out' phase chemical shift sagittal MRI sequences (rater 1 = 0.699, rater 2 = 0.718), as well as good agreement between raters for both 'in' and 'out' phase chemical shift sagittal MRI sequences (in phase = 0.656, 'out' phase = 0.576). However, when compared to the gold standard rating on a T2 based sequence, ratings on in' and 'out' phase MRI sequences overestimated the degree of stenosis. When comparing 'in' and 'out' ratings to the T2 ratings, agreement was poor with kappa ranging from 0.132 to 0.202.

Conclusion

Though both 'in' and 'out' phase chemical shift sagittal MRI sequences can be used to analyse neural foraminal stenosis, given its propensity to over-estimate the degree of stenosis in comparison to the T2 based images, assessment of the condition on these complementary limited sequences technique should be avoided/should be undertaken with caution.

Imaging the Posterior Elements of the Spine

The posterior elements of the spine consist of the pedicles, laminae, facets (articular processes), transverse processes, and the spinous process. They are essential for spinal stability, protecting the spinal cord and nerve roots, and enabling movement of the spine. Pathologies affecting the posterior elements can cause significant pain and disability. Imaging techniques, such as conventional radiography, computed tomography, and magnetic resonance imaging, are crucial for the diagnosis and evaluation of pathology, enabling accurate localization, characterization, and staging of the disease.